Add optional CH4 effects

[irm-codebase]

Fixes #18 and #5

This is my attempt to add more flexibility in how CH4 is treated.
Namely:

• Options on how to treat indirect and direct CH4 effects
  • "none": only direct
  • "all": everything (CO2 decay, indirect). Default, matches Watanabe's paper.
  • "no_carbon_cycle": useful for biogenic processes.

Other extra goodies:

• Safer imports (only for source code files I touched).
• Stronger tests.
• Small helper for 'fixed' RE (to reduce redundant code)

For disclosure:

• I am not an expert in atmospheric chemistry! I just did my best to adapt the code in Watanabe's SI.
• The code in the SI was a bit obtuse in some parts, so I did rely on an AI agent to help me parse / debug the code for some functions (namely while writing _agwp_ch4_cumulative). I've triple checked things, so it should be clean. However, the SI Code n.4 section full of 'magic values'. I'm trusting Watanabe's implementation, but it'd be ideal if these were converted to global constants with more useful names.
• NO2 code was left untouched (i.e., no indirect effect support). I tried to ensure the naming and approach would support this in the long run, if it is ever included.

[irm-codebase]

Some results for 2050, based on this figure from the paper:

Cum RF for SSP2 2.6 - MESSAGE (upper bound of GWP100 increase in Watanabe's results).

Cumulative RF for SSP4 8.5 - GCAM4 (almost the same as IPCC AR6)

[Copilot]

Pull request overview

Adds configurability for methane (CH4) prospective characterization by introducing selectable indirect-effects treatments, and updates validation tests to compare against Watanabe SI reference tables more directly.

Changes:

• Introduce indirect_effects modes for CH4 AGWP (e.g., "none", "no_carbon_cycle", "all") and thread the option through prospective CH4 RF + top-level characterize(...) for relevant metrics.
• Refactor prospective characterize_ch4 to reuse a centralized cumulative AGWP helper (_agwp_ch4_cumulative).
• Strengthen/expand pGWP100 tests for CH4 across years and scenarios, and add a monotonicity check across modes.

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 5 comments.

File	Description
dynamic_characterization/prospective/agwp.py	Adds CH4 indirect-effects modes and a shared cumulative AGWP implementation used by CH4.
dynamic_characterization/prospective/radiative_forcing.py	Exposes indirect_effects in prospective CH4 forcing and delegates to AGWP helper.
dynamic_characterization/dynamic_characterization.py	Adds indirect_effects parameter to the public API and forwards it where supported.
tests/test_prospective.py	Updates CH4 pGWP100 tests to match SI tables across scenarios/years; adds monotonicity test for CH4 modes.

Comments suppressed due to low confidence (1)

dynamic_characterization/prospective/radiative_forcing.py:25

• load_irf_ch4 and load_re_ch4 are imported but no longer used after switching characterize_ch4 to _agwp_ch4_cumulative. With flake8 enabled, this will raise an unused-import error; remove these imports (and any other now-unused CH4-specific constants/imports in this module).

from dynamic_characterization.prospective.agwp import _agwp_ch4_cumulative
from dynamic_characterization.prospective.config import get_scenario
from dynamic_characterization.prospective.data_loader import (
    load_irf_ch4,
    load_irf_co2,
    load_irf_n2o,
    load_re_ch4,
    load_re_co2,
    load_re_n2o,
)

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.